Skip to content

Use less GPU memory in test_managed_alloc_driver_undersubscribe. #188

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
bdice merged 2 commits into from
Apr 8, 2025
Merged

Use less GPU memory in test_managed_alloc_driver_undersubscribe. #188

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
5032137
Test h100 GPUs.
bdice Apr 2, 2025
4e0fab7
limit the size of memory allocated
isVoid Apr 7, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 5 additions & 5 deletions .github/workflows/conda-python-tests.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -72,19 +72,19 @@ jobs:
pull-request:
# amd64
- { ARCH: 'amd64', PY_VER: '3.9', CUDA_VER: '11.8.0', LINUX_VER: 'rockylinux8', GPU: 'l4', DRIVER: 'earliest', DEPENDENCIES: 'oldest' }
# TODO: Use CUDA 12.8.0 and h100 below.
- { ARCH: 'amd64', PY_VER: '3.13', CUDA_VER: '12.5.1', LINUX_VER: 'ubuntu24.04', GPU: 'l4', DRIVER: 'latest', DEPENDENCIES: 'latest' }
# TODO: Use CUDA 12.8.0 below.
- { ARCH: 'amd64', PY_VER: '3.13', CUDA_VER: '12.5.1', LINUX_VER: 'ubuntu24.04', GPU: 'h100', DRIVER: 'latest', DEPENDENCIES: 'latest' }
# arm64
- { ARCH: 'arm64', PY_VER: '3.11', CUDA_VER: '12.0.1', LINUX_VER: 'ubuntu20.04', GPU: 'a100', DRIVER: 'latest', DEPENDENCIES: 'latest' }
nightly:
# amd64
# TODO: Use CUDA 11.4.3 below.
- { ARCH: 'amd64', PY_VER: '3.9', CUDA_VER: '11.8.0', LINUX_VER: 'rockylinux8', GPU: 'l4', DRIVER: 'earliest', DEPENDENCIES: 'oldest' }
- { ARCH: 'amd64', PY_VER: '3.10', CUDA_VER: '11.8.0', LINUX_VER: 'ubuntu20.04', GPU: 'l4', DRIVER: 'latest', DEPENDENCIES: 'latest' }
- { ARCH: 'amd64', PY_VER: '3.10', CUDA_VER: '11.8.0', LINUX_VER: 'ubuntu20.04', GPU: 'h100', DRIVER: 'latest', DEPENDENCIES: 'latest' }
- { ARCH: 'amd64', PY_VER: '3.11', CUDA_VER: '12.0.1', LINUX_VER: 'rockylinux8', GPU: 'l4', DRIVER: 'latest', DEPENDENCIES: 'latest' }
- { ARCH: 'amd64', PY_VER: '3.12', CUDA_VER: '12.2.2', LINUX_VER: 'ubuntu22.04', GPU: 'l4', DRIVER: 'latest', DEPENDENCIES: 'latest' }
# TODO: Use CUDA 12.8.0 and h100 below.
- { ARCH: 'amd64', PY_VER: '3.13', CUDA_VER: '12.5.1', LINUX_VER: 'ubuntu24.04', GPU: 'l4', DRIVER: 'latest', DEPENDENCIES: 'latest' }
# TODO: Use CUDA 12.8.0 below.
- { ARCH: 'amd64', PY_VER: '3.13', CUDA_VER: '12.5.1', LINUX_VER: 'ubuntu24.04', GPU: 'h100', DRIVER: 'latest', DEPENDENCIES: 'latest' }
# arm64
# TODO: Use CUDA 11.4.3 below.
- { ARCH: 'arm64', PY_VER: '3.9', CUDA_VER: '11.8.0', LINUX_VER: 'ubuntu20.04', GPU: 'a100', DRIVER: 'latest', DEPENDENCIES: 'oldest' }
Expand Down
4 changes: 3 additions & 1 deletion numba_cuda/numba/cuda/tests/cudadrv/test_managed_alloc.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,7 +48,9 @@ def skip_if_cc_major_lt(self, min_required, reason):
def test_managed_alloc_driver_undersubscribe(self):
msg = "Managed memory unsupported prior to CC 3.0"
self.skip_if_cc_major_lt(3, msg)
self._test_managed_alloc_driver(0.5)
# We keep the allocation small so that it doesn't hang on GPUs
# with large memory (H100)
self._test_managed_alloc_driver(0.1)

# This test is skipped by default because it is easy to hang the machine
# for a very long time or get OOM killed if the GPU memory size is >50% of
Expand Down